AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

Density of States Prediction of Crystalline Materials via Prompt-guided Multi-Modal Transformer

Namkyeong Lee, Heewoong Noh, Sungwon Kim, Dongmin Hyun, Gyoung S. Na, Chanyoung Park

NeurIPS 2023 · Vision-language and multimodal › Human and embodied modeling · approach: Medical vision-language models · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add energy embeddings to E3NNE3NN · Phonon DOSMSE: 0.210 -> 0.200Improved
add energy embeddings to Graph NetworkGraph Network · Electron DOSMSE: 0.264 -> 0.240Improved
add energy embeddings to MLPMLP · Phonon DOSMSE: 0.309 -> 0.251Improved
add energy embeddings to E3NNE3NN · Electron DOSMSE: 0.296 -> 0.291No clear effect
add energy embeddings to E3NNE3NN · Band gap from predicted DOSMSE: 0.896 -> 1.605Hurt
add energy embeddings to Graph Network in OOD atom speciesGraph Network · OOD # atom speciesMSE: 0.460 -> 0.455No clear effect
add energy embeddings to Graph Network in OOD crystal systemGraph Network · OOD crystal systemR2: 0.505 -> 0.534Improved
DOSTransformer vs best baseline with energy (Electron DOS)DOSTransformer · Electron DOSMSE: 0.240 -> 0.221Improved
fine-tune only prompts instead of all parameters (ablation)DOSTransformer · OOD crystal system fine-tuningMSE: 0.365 -> 0.355Improved
remove either global or system loss (ablation)DOSTransformer · in-distributionMSEHurt
inject crystal system as one-hot into atom features instead of prompt before self-attention (ablation)DOSTransformer · Electron DOSMSE: 0.221 -> 0.222Hurt
inject prompt into atom features instead of before self-attention (ablation)DOSTransformer · Electron DOSMSE: 0.221 -> 0.227Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy